Margaret A.WinkerMD, Senior EditorIndividualAuthorPhil B.FontanarosaMD, Senior EditorIndividualAuthor
To the Editor.—The articles1,2 on the quality of POL testing are poor evaluations of the effects of CLIA. Dr Stull and colleagues1 state that 90% of laboratory errors occur in the preanalytic and postanalytic testing phases, yet the study only examines errors in PT, an artificial process used to detect problems in the analytic phase. It seems peculiar to devote so much attention to correcting problems in the test process least associated with laboratory errors. Given the steep learning curve associated with PT, it seems strange that the study only evaluated PT data from 1994, the first year of required PT for POLs. Error rates were counted by the number of "unsatisfactory PT events," yet CLIA considers "unsuccessful PT performance" (2 of 3 failed events) as significant. It appears that the authors are "proving" inferior testing in POLs with PT data from the earliest year available and using a more stringent definition of failure than that of the Health Care Financing Administration.
Rappo PD. Accuracy of Physicians' Office Laboratory Results. JAMA. 1998;280(2):129-132. doi:10-1001/pubs.JAMA-ISSN-0098-7484-280-2-jac80010